J Comp Physiol B

DOI 10.1007/s00360-008-0332-2

ORIGINAL PAPER

Food utilisation and digestive ability of aquatic and semi-terrestrial crayWshes, Cherax destructor and Engaeus sericatus (Astacidae, Parastacidae)

Stuart Maxwell Linton · Benjamin J. Allardyce · Wilhelm Hagen · Petra Wencke · Reinhard Saborowski

Received: 12 August 2008 / Revised: 15 December 2008 / Accepted: 16 December 2008© Springer-Verlag 2009

Abstract Both Engaeus sericatus and Cherax destructorare omnivorous crayWshes consuming a variety of fooditems. Materials identiWed in the faeces of both E. sericatusand C. destructor consisted of mainly plant material withminor amounts of arthropod animals, algae and fungi. Themorphology of the gastric mill of C. destructor suggeststhat it is mainly involved in crushing of food material whilethe gastric mill of E. sericatus appears to be better suited tocutting of food material. Given this, the gastric mill ofE. sericatus may be better able to cut the cellulose andhemicellulose Wbres associated with Wbrous plant material.In contrast, the gastric mill of C. destructor appears to bemore eYcient in grinding soft materials such as animal pro-tein and algae. Both species accumulated high amounts oflipids in their midgut glands (about 60% of the dry mass)which were dominated by triacylglycerols (81–82% of totallipids). The dominating fatty acids were 16:0, 16:1(n-7),18:1(n-9), 18:2(n-6), and 18:3(n-3). The two latter fatty

acids can only be synthesised by plants, and are thus indica-tive of the consumption of terrestrial plants by thecrayWshes. The similarity analysis of the fatty acid patternsshowed three distinct clusters of plants and each of thecrayWsh species. The complement of digestive enzymes,proteinases, total cellulase, endo-�-1,4-glucanase, �-gluco-sidase, laminarinase and xylanase within midgut gland sug-gests that both C. destructor and E. sericatus are capable ofhydrolysing a variety of substrates associated with anomnivorous diet. Higher activities of total cellulase, endo-�-1,4-glucanase and �-glucosidase indicate that E. sericatusis better able to hydrolyse cellulose within plant materialthan C. destructor. In contrast to E. sericatus, higher totalprotease and N-acetyl-�-D-glucosaminidase activity in themidgut gland of C. destructor suggests that this species isbetter able to digest animal materials in the form of arthro-pods. DiVerences in total cellulase and gastric millmorphology suggest that E. sericatus is more eYcient atdigesting plant material than C. destructor. However, thecontents of faecal pellets and the fatty acid compositionsseem to indicate that both species opportunistically feed onthe most abundant and easily accessible food items.

Keywords Engaeus sericatus · Cherax destructor · CrayWsh · Feeding · Nutrition · Stomach content · Gastric mill · Midgut gland · Storage lipids · Fatty acids · Digestive enzymes

Introduction

Crustaceans have successfully adapted to aquatic environ-ments. In the oceans, they represent the most abundant andthe most diverse phylum inhabiting pelagic as well asbenthic systems. They often appear in high numbers from

Communicated by I. D. Hume.

S. M. Linton (&) · B. J. AllardyceSchool of Life and Environmental Sciences, Deakin University, Pigdons Road, Geelong, VIC 3217, Australiae-mail: slinton@deakin.edu.au

W. Hagen · P. WenckeMarine Zoology, University of Bremen, P.O. Box 330 440, 28334 Bremen, Germany

R. SaborowskiAlfred Wegener Institute for Polar and Marine Research, Biologische Anstalt Helgoland, P.O. Box 180, 27483 Helgoland, Germany

123

J Comp Physiol B

the surface layers, down to deep-sea canyons, and thermalvents. Most of the members of crustacean taxa haveadapted to life in freshwater systems as well. However,only a number of species from a few taxonomic groups(e.g. isopods, anomurans and brachyurans) have success-fully invaded land.

The colonisation of land requires the adoption of a ter-restrial diet. For omnivorous/detritivorous species, thisinvolves consuming terrestrial plant material such asgrasses, leaf litter, fruits and seeds. There are a number ofdisadvantages associated with the consumption of plantmaterial. Notably, it is tough and Wbrous due to thepresence of large amounts of cellulose and hemicelluloseassociated with its cell walls. This makes plant materialslow and diYcult to digest (Linton and Greenaway 2007).Also given the low amounts of protein, plant material is ofpoor nutritive quality. To be able to utilise plant material asa nutrient source, decapod crustaceans require a gastricmill, this initially reduces the cellulose and hemicelluloseWbres into small particles. These particles are then enzymat-ically attacked by specialised enzymes such as cellulasesand hemicellulases, which hydrolyse them to their compo-nent sugars (Linton and Greenaway 2007).

Elucidating whether terrestrial digestive adaptationsoccur requires examination of a group of closely relatedcrustaceans that contain both aquatic and terrestrial species.The Australian parastacid crayWshes represent such a group(Crandall et al. 1999); members may be omnivorousaquatic species such as Cherax destructor or semi terres-trial burrowing species such as Engaeus sericatus.C. destructor is an aquatic freshwater crayWsh, commonlyknown as a yabby, and is found living in rivers, ponds,dams, and billabongs throughout eastern and centralAustralia (Merrick 1993; Jones and Morgan 1994). Yabbiesare considered omnivorous, feeding on a range of plant,algae, animal and detrital material (Goddard 1988;Faragher 1983; Beatty 2006). Burrowing parastacidcrayWsh, such those of the genus Engaeus are endemic tosoutheastern Australia and Tasmania. Engaeus are a semi-terrestrial genus, they require access to a permanent sourceof water but also engage in signiWcant terrestrial activityduring rain or at night to forage or mate (Horwitz 1990;Growns and Richardson 1988; Suter and Richardson 1977).They inhabit swampy areas and build burrows with charac-teristic chimneys (Horwitz 1990; Growns and Richardson1988). Their burrows may consist of either a single burrowor a complex of many burrows with interconnectingtunnels. At least one of these burrows extends down to thewater table or contains water. Some of the burrows termi-nate as feeding chambers under the roots of plants. As agroup, Engaeus species are omnivorous feeding on plantsroots, other plant material, and arthropods (Growns andRichardson 1988; Suter and Richardson 1977).

The functional morphology of the gastric mill, the com-position of storage lipids, as well as the activities of the pre-dominantly expressed digestive enzymes correlate wellwith dietary preferences (e.g. Heinzel 1988; Salindeho andJohnston 2003; Dalsgaard et al. 2003; Linton and Greenaway2004). The crustacean gastric mill is a part of the cardiacstomach. It consists of two lateral teeth and one medialtooth. In general, macrophagus crabs that consume largefood items such as animals, macro-algae and plants haverobust, dentate, heavily calciWed gastric mills. In contrast,microphagus/detritivorous crabs, which consume soft foodmaterial such as decayed seagrass, epiphytic algae and bac-teria, have less calciWed gastric mills which may possessstiV setae instead of being dentate (Kunze and Anderson1979; Heeren and Mitchell 1997; Salindeho and Johnston2003; Martin et al. 1998). Within the macrophagus crabs,the morphology of the medial and lateral teeth correlateswell with diet. The teeth of the gastric mill from the herbiv-orous sesarmid crab, Neosarmatium smithii possesses lowheavy transverse dentate ridges that interlock for cuttingtough Wbrous plant material (Giddins et al. 1986). The lat-eral teeth of omnivorous crabs such as Nectocarcinustuberculosus possess cusps for cutting food material andvertical ridges, which interact with surfaces on the medialtooth for grinding (Salindeho and Johnston 2003). Thus, thegastric mills of such omnivorous species appear to beadapted for mastication of both soft and hard food material.Lateral teeth from the gastric mill of carnivorous crabs suchas Ozius truncatus have large Xattened molar processes,which grind against robust ridges on the medial tooth forpulverising the relatively soft animal material (Skilleter andAnderson 1986). Although it has not been examined, thegastric mill of terrestrial omnivorous species is likely topossess morphological adaptations, such as heavily dentateteeth with low transverse ridges, for grinding and cutting ofthis tough Wbrous material.

Lipids are major energy storage products in crustaceans.They accumulate in the midgut glands and potentially reX-ect the fatty acid composition of the preferred food items,given the ingested fatty acids may be incorporated in stor-age triacylglycerols without biochemical modiWcation bythe organism (Nelson and Cox 2005). Furthermore, thepresence of fatty acids, such as poly unsaturated ones,which are known to be synthesised by plants and notanimals, may provide valuable information about theimportance of dietary plant material during the period inwhich the storage of lipids took place (Nelson and Cox2005). Some fatty acids may even serve as speciWc trophicmarkers as shown in marine environments (reviewed byDalsgaard et al. 2003).

The biochemical degradation of food items is facilitatedby a set of highly active digestive enzymes, which are syn-thesized in the midgut gland but subsequently, accumulate

123

J Comp Physiol B

in the stomach. Carnivorous species may produce highlevels of proteases such as trypsin and chymotrypsin fordigesting protein; omnivorous species express proteasesand carbohydrases such as �-amylase and �-glucosidase;algal and phytoplankton feeders have high levels of lamina-rinases for digesting laminarin, the major storage polysac-charide of algae; and herbivorous species show highactivities of cellulase and hemicellulase enzymes for hydro-lysing cellulose and hemicellulose found in plant cell walls(Johnston and Yellowlees 1998; Linton and Greenaway2004; Johnston and Freeman 2005; Crawford et al. 2005).Whilst the complement of digestive enzymes produced byterrestrial omnivorous species is largely unknown, it ismost likely to be dominated by high activities of cellulaseand hemicellulase enzymes. These enzymes may berequired to digest the terrestrial plant material. Indeed, interrestrial species, the cellulase activity may be higherwhile the laminarinase activity may be lower than that ofclosely related aquatic species, given the former may con-sume more plant material and fewer algae.

Little is known about the biochemical and morphologi-cal adaptations that Engaeus may possess towards anomnivorous diet. Given that they essentially forage in a ter-restrial environment, they may possess adaptations such asthose hypothesised above which enable them to eYcientlydigest terrestrial plant material. To elucidate dietary prefer-ence and biochemical and morphological adaptations thatEngaeus may possess, we examined the enzyme and lipidcomplement and morphology of the gastric mill of E. seric-atus. E. sericatus is endemic to ephemeral creeks south andsouthwest of the Otway ranges in central Victoria, Australia(Horwitz 1990). Possible adaptations to a terrestrial plantdiet were elucidated by comparing the enzyme comple-ment, activity and morphology of the gastric mill to that ofa close aquatic relative, C. destructor. The importance ofterrestrial plant material in the diet was assessed by exam-ining the presence of signature lipids, produced by plants,in the midgut glands of both species.

Materials and methods

Origin of samples

CrayWshes, C. destructor and E. sericatus, were caught inNovember 2006 in Victoria, Australia. C. destructor werecaptured from the ponds of Deakin University (WaurnPonds Campus, Geelong) with Jarvis Walker yabby traps.E. sericatus were collected from Birragurra creek atBirragurra. They were dug out of muddy soil, roughlycleaned and immediately transferred to the laboratories ofDeakin University while being kept chilled. In the labora-tory, they were placed into separate tubes or beakers. The

animals were left in these facilities overnight to defecate.The next day, the animals were removed and the faeceswere collected using forceps and preserved in 70% alcohol.Pieces of the most abundant plants from the sampling sites,including roots, stalks, and leaves were taken to the labora-tory and were deep-frozen for later fatty acid analysis.

Dissection of the digestive organs

Prior to dissection, the animals were anaesthetized by cool-ing them for 30 min on ice. The lengths of the animals weremeasured from the tip of the rostrum to the end of the telsonand the fresh weights were determined. The carapace wasthen opened by lateral incisions with scissors and was care-fully detached dorsally from the inner organs. The midgutgland tissue was completely removed and weighed. Themass of the midgut gland was calculated in relation to thetotal body mass of the animal providing the midgut glandindex (MGI). An aliquot of tissue was immediately used toprepare extracts for cellulase, hemicellulase and chitinaseassays while another aliquot was deep-frozen in liquidnitrogen for later proteinase assays and lipid determina-tions. The stomach was completely removed and stored in70% ethanol for later scanning electron microscopy.

Anatomy of gastric mill

Cardiac stomachs of C. destructor and E. sericatus contain-ing the gastric mills were carefully dissected out of theanimal. Connective tissue surrounding the stomachs wasremoved to reveal the gastric mill by careful dissection andby soaking the stomachs in 5% sodium hydroxide for2–3 h. Gastric mills were then rinsed with distilled water,dried in graded ethanol (80, 90 and 100%) and air driedovernight at room temperature. Dried samples were thenmounted on an scanning electron microscope stub usingdouble sided carbon tape (Spi carbon tape, 5072) and sput-ter coated with gold using a sputter coater. Gold-coatedsamples of the gastric mills were examined using a PhillipsXL20 scanning electron microscope.

Analysis of food items in the faeces

Food items within the faeces were identiWed using lightmicroscopy. The faecal samples were smeared onto a glassslide, stained with 0.5% (w/v) Congo red and examinedunder a compound light microscope. Plant items were iden-tiWed by the presence of cellulose Wbres, water vascularbundles, or cells with thick cell walls. Animal material wasidentiWed by the fragments of chitinous arthropod append-ages. Algal Wlaments were identiWed by chains of emptycells with thick cell walls. Hyphae containing sporocystswere deemed as fungi.

123

J Comp Physiol B

Lipid and fatty acid analysis

Subsamples of midgut gland tissue for later lipid and fattyacid analysis were transferred at ¡198°C to Germany. Thesamples were lyophilised for 48 h and weighed. Total lipidswere extracted with dichloromethane/methanol (2:1 v/v)and determined gravimetrically (Hagen 2000). Lipid clas-ses were separated and analysed by thin-layer chromatogra-phy Xame ionisation detection (TLC-FID) with anIATROSCAN analyser (Iatron Laboratories, Inc., MK-5TLC/FID Analyser) as per Fraser et al. (1985). Theextracted lipids were hydrolysed and the fatty acids wereconverted to methyl ester derivatives (FAMEs). The con-version was carried out for 4 h at 80°C in a solution ofmethanol and 3% (v/v) sulphuric acid (Kattner and Fricke1986). After cooling to room temperature, 4 ml of deion-ised water was added. FAMEs were extracted with hexane(3 £ 1.7 ml) and analysed in a gas chromatograph (HewlettPackard Model No 6890A). The device was equipped witha DB-FFAP column (30 m length, 0.25 mm inner diameter,0.25 �m Wlm thickness). The separation was performed byusing temperature programming and helium as carrier gas.FAMEs were detected by Xame ionisation and identiWed bycomparing retention time data with those obtained fromstandard mixtures.

Enzyme assays

Extracts for enzyme assays were homogenised in0.1 mol l¡1 sodium-acetate buVer (pH 5.5) containing2 mmol l¡1 dithioerythritol as a reducing reagent or indemineralised water. Depending on their mass, the midgutgland samples (50–1,000 mg) were homogenised in 1–3 mlof liquid using a Polytron homogenizer or an ultrasonic celldisruptor (Branson, SoniWer). The homogenates were cen-trifuged for 10 min at 10,000g and at 4°C. The supernatants(extracts) were aliquoted and stored frozen until enzymeanalysis.

Assays for cellulase [total cellulase, endo-�-1,4-glucan-ase (EC 3.2.1.4) and �-glucosidase (EC 3.2.1.21)] andhemicellulase (laminarinase [endo-�-1,3-glucanase (EC3.2.1.39)], licheninase [endo-�-1,3; 1,4 glucanase (EC3.2.1.73)], xylanase [endo-�-1,4-xylanase (EC 3.2.1.8)])were carried out as described by Linton and Greenaway(2004). Total cellulase activity was measured in 50 �l ali-quots of enzyme extract. Aliquots of 25 �l of the enzymeextract were used to measure the activities of �-glucosi-dase, laminarinase, xylanase and lichenase. Ten microlitrealiquots of the enzyme extract were used in the endo-�-1,4-glucanase assays. All enzyme assays were incubated at30°C. Total cellulase and �-1,4-glucosidase (E.C. 3.2.1.21)activities were measured as the rate of glucose productionfrom microcrystalline cellulose (Sigmacell 20, Sigma S-3504)

and cellobiose (Sigma, C-7252), respectively. Endo-�-1,4-glucanase (E.C. 3.2.1.4), laminarinase (EC 3.2.1.39),lichenase (EC 3.2.1. 73) and xylanase (EC 3.2.1.8) activi-ties were measured as the rate of reducing sugar productionfrom the hydrolysis of the respective substrates, carboxy-methyl cellulose (Sigma, C-5678), laminarin (from Lami-naria digitata; Sigma, L-9634), lichenan (from Cetrariaislandica; Sigma L-6133) and xylan (from birchwood,Betula; Sigma X-0502).

Chitinase [1,4-�-poly-N-acetylglucosaminidase (EC3.2.1.14)] and NAGase [N-acetyl-�-D-glucosaminaidase(EC 3.2.1.52)] were assayed as per Saborowski et al.(1993). Chitinase activity was measured as the release ofRemazol Brillant blue from the substrate, CM-Chitin-RBB(Loewe Biochemica, 04106). NAGase activity was deter-mined by the release of p-nitrophenol from p-nitrophenyl-N-acetyl-�-D-glucosaminide [NAGpnp, (Sigma N-9376)].

The assays for total proteinase, trypsin-like and chymo-trypsin-like activities were as outlined by Saborowski et al.(2004). Total proteolytic activity was determined by therelease of azo dye from the hydrolysis of azocasein (Fluka,11615). Trypsin-like, chymotrypsin-like and alanine-amin-opetidase enzyme activities (E.C. 3.2.1) were determinedby the release of p-nitroaniline from the substrates N�-Ben-zoyl-L-arginine-4-nitroanilide-hydrochloride (L-BAPNA,Fluka 12915) (Erlanger et al. 1961), N-Succinyl-Ala-Ala-Pro-Phe-p-nitroanilide (SAAPPNA, Sigma S7388) andL-Alanine-p-nitroanilide-hydrochloride (Sigma, A 9325).

Inhibition assays of proteinases were carried out afterTeschke and Saborowski (2005) with the cysteine protein-ases inhibitor Trans-Epoxy-Succinyl-L-Leucylamido-(4-Guanidino)-Butane (E64, Sigma E 3132) and the serineproteinase inhibitor 4-(2-Aminoethyl)enzenesulfonyl-Xuo-rid Hydrochloride (AEBSF, Merck 124839). The inhibitedsamples were subjected to proteinase assays with azocaseinas substrate.

Protein concentrations in the enzyme extracts weredetermined after Bradford (1976) using a commercial dyereagent (BioRad 500-0006). Bovine �-globuline (MP Bio-chemicals 11BSAG010) at concentrations of 0–1 mg ml¡1

served as standards.

Statistics

Statistical analyses were carried out with the computerprograms STATISTICA Ver 5.5 (Stats soft) orSigmaStat Ver 3.1 (Systat Software). Tests for normaldistribution and equal variances were run initially.Pairwise comparison of data sets was performed usingeither Student’s t test or a Mann–Whitney rank sum test.Means were deemed to be signiWcantly diVerent if thecalculated probability of the statistical test was less than0.05.

123

J Comp Physiol B

Results

Morphometric data

The Cherax-specimens were generally larger than theEngaeus-specimens and had a wider mass range (Table 1).The sex distribution was similar in both species (25–50%females). The masses of the midgut glands increased allo-metrically with the body masses [E. sericatus: Midgutgland mass (g) = 0.0562 £ Animal mass (g) + 0.0249(r2 = 0.9659), C. destructor: Midgut gland mass (g) =0.028 £ Animal mass (g) + 0.5742 (r2 = 0.8129)]. Twoindividuals of C. destructor showed exceptionally smallmidgut glands: one of them was a gravid female with adepleted midgut gland while the reason for the seconddeviating animal, a male, remained unclear. The midgutgland indices (MDI) of E. seratus did not vary signiWcantlyfrom those of small C. destructor. However, a signiWcantdiVerence appeared between E. seratus and largeC. destructor as well as between small and large C. destructor(Table 1).

Items identiWed in the faeces

Material identiWed in the faeces suggested that E. sericatuswas omnivorous, consuming mainly plant material withminor amounts invertebrates, algae, fungi and invertebrates(Table 2). Of the materials identiWed, the majority consistedof plant material (leaf fragments with stomata, cellulosebundles constituting the water vascular system and clustersof cells with thick cell walls), with minor amounts of fungi(hyphae with sporocysts) and animal material (fragments ofchitinous invertebrate appendages) (Table 2). Substantialamounts of the matter in the feaces could not be identiWed.

Like E. sericatus, C. destructor was an omnivore con-suming mainly plant material with signiWcant amounts ofWlamentous algae and minor amounts of arthropods andfungi (Table 2). The majority of the food material identiWedin the faeces consisted of plant material in the form of

macerated leaf material and bundles of cellulose Wbrescomprising the water vascular system. Minor amounts ofanimal and fungal material were also identiWed in thefaeces. The faeces of both C. destructor and E. sericatuscontained similar percentages of plant, arthropod andfungal material. Algal Wlaments in the form of long chainsof empty cells, most likely Spirogyra, made up approxi-mately 25% of the identiWable items in the faeces ofC. destructor. This percentage was signiWcantly higher thanthat in than E. sericatus (Table 2).

Anatomy of gastric mills

Engaeus sericatus

The medial tooth was part of the urocardiac ossicle(Fig. 1a). The posterior part of the medial tooth protrudedventrally, it curved round and ended in two sharp ventro-lateral cusps, which pointed anteriorly. On the lateral sidesof the protrusion were deep indentations. Anterior to theprotrusion were two lateral ridges, the ends of which termi-nated as sharp anterior pointing cusps. The urocardiac ossi-cle immediately anterior to the dentate region of the medialtooth sloped away dorsally.

The lateral tooth was dentate (Fig. 1b). The ventral sur-face of the lateral tooth possessed eight large curved cusps,which pointed posteriorly. Anterior cusps were larger thanthe posterior ones, with the size progressively decreasingfrom anterior to posterior. The largest of the cusps wereapproximately 250 �m long and 200 �m wide anddecreased in size to approximately 58 �m. These cuspsgave the ventral surface of the lateral tooth a serratedappearance. The cusps from anterior to posterior were pro-gressively rotated so that the last posterior cusp was 90° tothe Wrst anterior cusp (Fig. 1b, c). Next to the Wrst cusp, wasa singular ventral cusp (Fig. 1b). Cusp number two was sin-gular. Next to and in line with the remaining six cusps weresmaller cusps (Fig. 1b). Dorsal to ventral cusps three andfour were smaller singular cusps. On the dorsal side of ven-tral cusps, numbered Wve to eight were bumpy ridges. Thedorsal surface of the lateral tooth displayed evidence ofwear in the form of scoring and indentations. The cusps on

Table 1 Morphometric data of specimens used in this study

Values are given as range between minimum and maximum or as meanvalues § SEM. For the midgut gland index, similar superscript lettersindicate similar mean values (1-way ANOVA, P ¸ 0.05)

Engaeus sericatus

Cherax destructor

Number of animals 8 6 (small) 5 (large)

Length (mm) 38–60 62–88 111–125

Sex distribution 3f, 5m 1f, 4m 2f, 4m

Fresh mass animal (g) 1.87–10.92 5.76–16.7 28.6–71.4

Mass of midgut gland (g) 0.11–0.61 0.32–1.08 1.65–2.66

Midgut gland index (%) 6.26 § 0.23a 5.89 § 0.42a 3.96 § 0.55b

Table 2 Percentages of identiWable items in the faeces [Mean § SEM(n)]

Within a row, similar superscript letters indicate similar means(P ¸ 0.05, t test)

Cherax destructor Engaeus sericatus

Arthropod fragments 3.3 § 0.8 (6)a 7.4 § 3.4 (5)a

Plant material 71.7 § 13.9 (6)a 89.6 § 3.1 (5)a

Algal Wlaments 24.3 § 13.1 (6)a 1.2 § 1.2 (5)b

Fungal hyphae 0.7 § 0.7 (6)a 1.8 § 1.8 (5)a

123

J Comp Physiol B

Fig. 1 Scanning electron micrographs of the gastric mills from Engaeus sericatus (a–d) and Cherax destructor (e–g). a Ventral view of the medial tooth from E. sericatus with ventro lateral cusps (VC), lateral indentations (I) and two lateral ridges (LR1 and LR2) which terminate as sharp anterior pointing cusps (asterisk). Anterior (Ant) posterior (Post) orientation of the medial tooth is also indicated. b Lateral tooth from the gastric mill of E. sericatus with 8 curved cusps (numbered 1–8), ventral cusp (VC) and bumpy ridges on the posterior part of the tooth (asterisk). Orientation of the tooth as in vivo is also indicated Anterior (Ant), Posterior (Post). c Lateral view of the medial tooth and right lateral tooth with lateral accessory tooth (arrow) and setal brush (asterisk). d Ventral view of the gastric mill of E. sericatus with medial (MT) and lateral teeth (LT). Anterior (Ant) posterior (Post) orientation of the gastric mill is also indicated. e Medial tooth from C. destructor with ventrally raised posterior (Post) portion with indentations on the lateral sides (I), laterally pointing cusps (asterisk), anterior ridge (R) and lateral ridge (LR) immediately anterior (Ant) of the ventral protrusion. File like surface of the indentation on the lateral side of the ventral protrusion is also indicated with an arrow. f Lateral tooth from the gastric mill of C. destructor with 4 distinct cusps [numbered 1–4 from the anterior portion (ant) of the tooth], ventral incisor like cusp (IC) and bumpy sole (S) of the posterior part of the tooth. g Ventral view of the gastric mill of C. destructor with medial tooth (MT), lateral teeth (LT), lateral accessory tooth (asterisk) and setal brush (arrow)

123

J Comp Physiol B

the lateral teeth of intermoult animals appeared to beslightly longer and sharper than those of premoult animals(not shown).

The lateral accessory tooth consisted of a large coneshaped cusp (t590 �m) which was ventral to the lateraltooth and pointed posteriorly (Fig. 1c). In between the lat-eral accessory tooth and the lateral teeth was a row of brushlike setae (Fig. 1c).

The lateral teeth were positioned parallel to the urocar-diac ossicle, which contained the medial tooth (Fig. 1d).The medial and lateral teeth were brown, indicating that thecuticle covering the teeth was possibly tanned. The serra-tions of the lateral teeth Wtted into the triangular indentationon the medial tooth (Fig. 1c). Anterior pointing cusps onthe medial tooth Wtted in between the ventral cusps andcusp one on the lateral tooth.

Cherax destructor

The posterior part of the urocardiac ossicle was dentate andformed the medial tooth. The posterior part of the medialtooth was raised ventrally and ended in two lateral pointingcusps (Fig. 1e). This is in contrast to that of E. sericatus,where the medial protrusion of the medial tooth terminatedas anterior pointing cusps (Fig. 1a, c). In between thesecusps was a V shaped valley that formed into an anteriorridge (Fig. 1e). On the lateral sides of the ventral protrusionwere two indentations. Their surfaces had a Wle like appear-ance. Anterior to the protrusion of the medial tooth was onerounded lateral ridge. The ridge was W shaped with araised central portion. Unlike the lateral ridges of themedial tooth of E. sericutus these ridges did not terminateas sharp lateral cusps (Fig. 1a, e). The urocardiac ossicleimmediately anterior to the dentate medial tooth slopedaway dorsally.

Four distinct rectangular cusps were present at the ante-rior end of the lateral tooth (Fig. 1f). These cusps wereinterconnected by an anterior–posterior ridge. In premoultanimals, these cusps were worn down to the level of thisanterior–posterior ridge. On the ventral side of cuspnumbered two and three was a cusp which superWciallyresembled an incisor. Posterior to these four cusps was aXat sole like structure, the surface of which was bumpy inintermoult animals. In contrast, this surface was worn Xatand displayed scoring marks that ran along the anterior–posterior axis in a premoult animal. The lateral tooth thusdisplayed obvious wear, which must have occurred overthe moult cycle. The structure of the lateral teeth ofC. destructor was considerably diVerent to that of E. seric-atus. The lateral teeth of E. sericatus possessed sharp tri-angular cusps for cutting while the cusps of the lateralteeth from C. destructor were rectangular with a sole likeposterior part (Fig. 1b, f).

The lateral accessory tooth consisted of three curved peglike cusps (Fig. 1g). These cusps were ventral to the lateraltooth and pointed posteriorly. Posterior to the lateral acces-sory tooth and ventral to the lateral tooth was a row ofbrush like setae.

The lateral teeth were parallel to the urocardiac ossiclecontaining the medial tooth (Fig. 1g). The sole like poster-ior part of the lateral teeth faced the ventral protrusion ofthe medial tooth. This sole like plate of the lateral teeth mayhave abutted against the Wle like surface of the most dorsalpart of the posterior protrusion of the medial tooth. Thismay have represented the grinding part of the gastric mill.

Like E. sericatus, the serrated dorsal edge of the anteriorpart of the lateral tooth may have Wtted into the lateral indenta-tions on the medial tooth (Fig. 1g). The ravines between thesecutting teeth were not as distinct as those of E. sericatus, giventhe presence of the anterior–posterior ridge. The anterior ven-tral incisor like cusp on the lateral teeth may have comple-mented the ventral lateral cusps of the protrusion of the medialtooth. The lateral teeth may have been able to rotate on theanterior–posterior axis to meet in the middle of the V shapedvalley between the lateral cusps of the medial tooth.

Lipid content and composition of lipid classes

The midgut gland of both E. sericatus and C. destructorcontained similar amounts of total lipid (Table 3). In eitherspecies, the lipids were dominated by triacylglycerols (81–82% of total lipids) which amounted to approximately 50%of the dry mass. Wax esters were present in the midgutgland of both species, but their concentration was higher inE. sericatus than in C. destructor (Table 3). Free fatty acidsand sterols were not detected in the midgut gland of eitherspecies. Polar lipids comprised 10–15% of total lipids andtheir levels were signiWcantly higher in the midgut gland ofC. destructor than in that of E. sericatus (Table 3).

Fatty acids and fatty alcohols

Of the three saturated fatty acids (14:0, 16:0, and 18:0) iden-tiWed, fatty acid 16:0 was the most predominant, comprising16 and 18% of total lipid in E. sericatus and C. destructorrespectively (Fig. 2a). The amounts of the fatty acids 14:0and 16:0 did not diVer signiWcantly between species whilethe amount of 18:0 was signiWcantly higher in the midgutgland of E. sericatus than in that of C. destructor.

Another major fatty acid was the monounsaturated fattyacid 18:1(�-9). Its levels in the midgut gland of both spe-cies were similar and it made up 13% of total fatty acids inC. destructor and 16% of total fatty acids in E. sericatus(Fig. 2a). However, the levels of two other monounsatu-rated fatty acids, 16:1(�-7) and 18:1(�7) were signiWcantlyhigher in the midgut gland of E. sericatus (Fig. 2a).

123

J Comp Physiol B

The polyunsaturated fatty acid (PUFA) 18:2(�-6) domi-nated the polyunsaturated fatty acids in C. destructor. Incontrast, the PUFAs from the midgut gland of E. sericatuswere not dominated by one particular fatty acid (Fig. 2a).The levels of the PUFAs, 18:2(�-6), 18:3(�-3), 20:2(�-6)

and 20:4(�-6) were higher in the midgut glands fromC. destructor than the midgut glands from E. sericatus. Incontrast, the amounts of the PUFAs 20:5(�-3) and 22:6(�-3)were signiWcantly higher in the midgut glands fromE. sericatus (Fig. 2a). UnidentiWable fatty acids were pres-ent in the midgut glands of both species. Their amount,however, was small and reached 3.5% in E. sericatus and5.7% in C. destructor (Fig. 2a). Fatty alcohols were notpresent in C. destructor, but traces of the fatty alcohols14:0A, 16:0A and 18:0A were present in the midgut glandof 4 out of 8 specimens of E. sericatus. However, thesevalues did not exceed 0.5% the total amount of fatty acids.

The dominating fatty acids from plant material were16:0, 18:2(�-6), and 18:3(�-3) amounting to 16–26% oftotal fatty acids (Fig. 2b). The monounsaturated fatty acid18:1(�-9) accounted for 6% of total fatty acids. Each of theother fatty acids amounted to less than 2% of the total fattyacids except the fraction of unidentiWable fatty acids, whichaccounted for 9% of all fatty acids.

The similarity analysis of the fatty acid yields (%)revealed three distinct clusters separating the plant materialsamples from the crayWsh midgut glands at the 70% simi-larity level and distinguishing between both crayWsh spe-cies at the 82% level (Fig. 3).

Enzyme activities

Total cellulase activity measures the production of glucosefrom the hydrolysis of cellulose. The midgut gland of bothE. sericatus and C. destructor possessed low activities oftotal cellulase (Fig. 4). Although variable the total cellulaseactivity in the midgut gland of E. sericatus was 2.7 timeshigher than that of C. destructor (Fig. 2). Given cellobiohy-drolase is purported to be absent in arthropods, the combinedaction of endo-�-1,4-glucanase and �-1,4-glucosidase may

Table 3 Total lipid concentration and composition of lipid classesfrom the midgut glands of Cherax destructor (n = 7) and Engaeussericatus (n = 8) in terms of percentage of dry mass (%DM) andpercentage of total lipids (%TL)

Mean values § SEM. Within a row, diVerent superscript lettersindicate statistically diVerent mean values (P < 0.05, t test)

NT not tested

Cherax destructor

Engaeus sericatus

Total lipids (TL)

%DM 56.4 § 4.3a 63.0 § 6.2a

Neutral lipids

Wax esters (WE)

%DM 1.9 § 0.8a 4.9 § 0.9b

%TL 3.3 § 1.5a 8.1 § 1.4b

Triacylglycerols (TAG)

%DM 49.2 § 3.3a 51.4 § 5.4a

%TL 81.6 § 2.1a 81.0 § 1.6a

Free fatty acids (FFA)

%DM 0 § 0 0 § 0 NT

%TL 0.1 § 0.1 0.1 § 0.1 NT

Sterols (ST)

%DM 0 § 0 0 § 0 NT

%TL 0 § 0 0.1 § 0.1 NT

Membrane lipids

Polar lipids (PL)

%DM 8.8 § 0.4a 6.7 § 0.7b

%TL 15.0 § 1.2a 10.7 § 0.7b

Fig. 2 Percentage composition of selected fatty acids (>1% of total fatty acids) from the midgut glands of the crayWsh, Cherax destructor and Engaeus sericatus (a) and from plant material (b) taken at the sampling sites. Mean values + SEM of n = 8, each. Levels of signiWcance between species are indicated by asterisks at P < 0.05*), P < 0.01% (**), and P < 0.001% (***). The grey bars emphasize the dominating fatty acids

14:0 14:0

18:1(n-9)

18:0

16:3(n-4)

16:1(n-7)

***

***

18:1(n-9)

18:0

16:3(n-4)

16:1(n-7)

Fat

ty a

cids

18:4(n-3)

18:3(n-3)

18:2(n-6)

18:1(n-7)

*

***

***

18:4(n-3)

18:3(n-3)

18:2(n-6)

18:1(n-7)

22:6(n-3)

20:5(n-3)

20:4(n-6)

20:2(n-6)

Cherax destructor

**

***

***

***

***

22:6(n-3)

20:5(n-3)

20:4(n-6)

20:2(n-6)

Amount (% of total fatty acids)

0 5 10 15 20 25

Amount (% of total fatty acids)

0 5 10 15 20 25 30 35

a) b)

Engaeus sericatus

16:0 16:0

unknown unknown

123

J Comp Physiol B

account for the total cellulase activity (Scrivener andSlaytor 1994; Watanabe et al. 1997). Consistent with this,the endo-�-1,4-glucanase and �-glucosidase activities inthe midgut gland of E. sericatus were, respectively, 2.3 and3.4 times higher than that of C. destructor (Fig. 4). Endo-�-1,4-glucanase activity was the highest of all of the cellulaseand hemicellulase enzymes measured.

The midgut gland of both E. sericatus and C. destructorpossess similar but substantial activities of the hemicellu-lase enzymes, laminarinase, xylanase and lichenase(Fig. 5). The order of decreasing enzyme activity waslaminarinase, lichenase and xylanase.

The midgut gland of both species contained similaractivities of endochitinase (Fig. 6). However, the activity ofN-acetyl-�-D-glucosaminidase was higher in the midgutgland of C. destructor than in that of E. sericatus.

The total protease activity in the midgut gland ofC. destructor was higher than that of E. sericatus (Fig. 7).In contrast, the midgut gland of E. sericatus containedhigher activities of chymotrypsin like protease and alanine-aminopeptidase than the midgut gland of C. destructor. Themidgut glands of both species showed similar levels oftrypsin like protease.

Proteinase classes

In the midgut gland of both species, the serine proteinasesdominated the proteinases given the serine proteinaseinhibitor AEBSF reduced total proteolytic activity by morethan 50% in C. destructor and 75% in E. sericatus (Fig. 8).Low levels of cysteine proteinase were present in themidgut glands of both species. The cysteine proteinaseinhibitor E64 reduced total proteolytic activity by 20% inC. destructor and 5% in E. sericatus. 20% and 15% of totalproteolytic activity remained, respectively, in E. sericatusand C. destructor after inhibition of both the serine andcysteine proteinases.

Fig. 3 Dendrogram of Bray-Curtis similarities between the fatty acidpatterns of plant material, and the midgut glands of the crayWshesE. sericatus and C. destructor

Fig. 4 Activities of total cellulase (�mol glucose min¡1 g¡1 tissue) (a), endo-�-1,4-glucanase (�mol reducing sugars produced min¡1 g¡1 tissue) (b) and �-glucosidase (�mol glucose min¡1 g¡1 tissue) (c) in the midgut glands of Cherax destructor (C. d.) and Engaeus sericatus (E. s.). The bars represent mean values + SEM of n = 6–9 animals. Asterisks indicate that the mean values between species were statistically diVerent (P < 0.05)

a) b)

c)

123

J Comp Physiol B

Discussion

In this study, we compared the feeding ecology and physiol-ogy of two closely related crayWsh species by means of thegastric mill morphology, gut contents, storage lipids and

fatty acid composition in the midgut gland, and the enzy-matic ability to cleave protein, chitin and plant structuralcompounds such as cellulose and hemicellulose. Theseresults revealed adaptations related to the utilisation of plantmaterial as a consequence of the colonisation of land.

Fig. 5 Activities of hemicellu-lose degrading enzymes, lamina-rinase (�mol reducing sugars produced min¡1 g¡1 tissue) (a), xylanase (�mol reducing sugars produced min¡1 g¡1 tissue) (b) and lichenase (�mol reducing sugars produced min¡1 g¡1 tis-sue) (c) in the midgut glands of Cherax destructor (C. d.) and Engaeus sericatus (E. s.). The bars represent mean values + SEM of n = 6–9 animals. Bars above the mean values indicate that the enzyme activities were statistically simi-lar between species (P ¸ 0.05)

a) b)

c)

Fig. 6 Activities of chitin degrading enzymes endo-chitinase (�absorbance 550 nm min¡1 g¡1 tissue) (a) and N-acetyl-�-D-glucosa-minidase (� absorbance 405 nm min¡1 g¡1 tissue) (b) in the midgutglands of Cherax destructor (C. d.) and Engaeus sericatus (E. s.). Thebars represent mean values + SEM of n = 6–9 animals. The bar above

the means for endo-chitinase activity indicate that it was statisticallysimilar between species (P > 0.05), while the asterisk indicates that theN-acetyl-�-D-glucosaminidase activity between species was statisti-cally diVerent (P < 0.05)

a) b)

123

J Comp Physiol B

Functional morphology of the gastric mill

Both C. destructor and E. sericatus have well-developedgastric mills, with both species possessing ossiWeddentate lateral and medial teeth. The morphology of thegastric mill of E. sericatus suggests that it is primarilyinvolved in cutting, while the morphology of the gastricmill of C. destructor suggests that it is primarilyinvolved in crushing.

The coordinated movements of the medial tooth andlateral teeth of the gastric mill shreds and pulverises foodmaterial. The movements of these teeth can be divided intothe cutting and grinding mode and the squeeze mode(Heinzel 1988; Heinzel et al. 1993). During the cutting andgrinding mode in the gastric mill of Cherax, the posterior solelike part of the lateral teeth would meet the Wle like part of themedial tooth (triangular indentation on the dorso- lateral sidesof the medial tooth (Fig. 1e). As the medial and lateral teeth

Fig. 7 Activities of total proteinase (�absorbance 366 nm min¡1 g¡1 tissue) (a) and protein degrading enzymes trypsin (�mol p-nitrophenol min¡1 g¡1 tissue) (b), chymo-trypsin (�mol p-nitrophenol min¡1 g¡1 tissue) (c) and alanine-aminopeptidase (�mol p-nitrophenol min¡1 g¡1 tissue) (d) in the midgut glands of Cherax destructor (C. d.) and Engaeus sericatus (E. s). The bars represent mean values + SEM of n = 6–9 animals. The asterisks indicate that the enzyme activities diVered signiWcantly between species (P < 0.05), while the bars indicate that the enzyme activity was similar between species (P > 0.05)

a) b)

d)c)

Fig. 8 EVects of cysteine [Trans-Epoxy-Succinyl-L-Leucylamido-(4-Guanidino)-Butane (E64)] and serine [4-(2-Aminoethyl)enzenesulfo-nyl-Xuorid Hydrochloride (AEBSF)] protease inhibitors on the activities of protein degrading enzymes in the midgut glands of Cherax destructor (a) and Engaeus sericatus (b). The bars represent mean values + SEM of n = 6 animals. Asterisks indicate that the mean values diVered signiWcantly between species (P < 0.05)

a) b)

123

J Comp Physiol B

move past each other food material is presumably crushedbetween them and rolled against the Wle like surface. Thisaction would most eYciently rupture cells and release thecytoplasmic contents of soft material such as animal materialand algae. At the end of the cutting and grinding mode, thefour lateral cusps of the anterior part of the lateral toothwould encounter the Wle like surface of the medial tooth ormay Wt into the triangular indentation on the medial tooth(Fig. 1g). This may cut the food material and would presum-ably be able to cut the tough plant material. In contrast, thelateral teeth of the gastric mill from E. sericatus possess anumber of serrations and hence are saw like (Fig. 1b, f).During the cutting and grinding action, these serrations of thelateral teeth would meet with the large lateral triangularindentation of the medial tooth. This cutting structure may bemost eYcient at cutting tough Wbrous, plant material.

The lateral accessory teeth are believed to aid in thetransfer of food material from the lumen of the cardiacchamber into the path of the teeth of the mill (Salindehoand Johnston 2003; Kunze and Anderson 1979). Unlike thebrush-like structures found in most brachyuran species(Kunze and Anderson 1979; Martin et al. 1998), the acces-sory teeth of both C. destructor and E. sericatus are sharp,well-calciWed coned shaped cusps. They are also positionedmore posteriorly than in the brachyurans, indicating theymay help to hold food in the mill during mastication ratherthan just feed material into it.

During the squeeze mode in C. destructor, the lateralteeth may move in and twisted on their anterior posterioraxis so that the anterior ventral cusps on the lateral teethmeet with the deep V shaped valley between the ventro-lat-eral projecting cusps of the medial tooth (Suthers andAnderson 1981; Heinzel 1988). During a similar action inE. sericatus, the anterior pointing lateral cusps of themedial tooth may also move into the valley between theanterior serrated edge and the anterior ventro-lateral cusp ofthe lateral tooth. This action, depending on the forceapplied by the medial and lateral teeth, would eithersqueeze or chop pieces of food.

The general morphology of the medial and lateral teethfrom the gastic mills of both C. destructor and E. sericatusis similar to that of other omnivorous species such asNectocarcinus tuberculosus. That is, the lateral teeth pos-sess cusps for cutting food material and vertical ridges forgrinding (Salindeho and Johnston 2003). The medial toothpossesses surfaces for grinding and structures such as ven-tral pointing cusps to aid in squeezing/cutting (Salindehoand Johnston 2003).

Faecal pellet contents

The diets of both C. destructor and E. sericatus (Table 2)were similar to that of other burrowing parastacid crayWsh

such as Engaeus cisternarius, E. fossor and Parastacoidestasmanicus tasmanicus which consists of mainly plantmaterial (roots of button grass, phloem and xylem fromfallen logs, leptosporangia of the treefern, Dicksonia antar-tica) with minor amounts algal, fungal and animal material(Suter and Richardson 1977; Growns and Richardson1988). Such burrowing parastacid crayWsh species aremainly subterranean feeders consuming plant roots, whichgrow into their burrows (Gowns and Richardson 1988), andit seems likely, that plant roots serve as a valuable and easyaccessible food source.

As observed in this and other studies, crayWsh of thegenus Cherax are also detritivores/omnivores (Faragher1983; Beatty 2006). They consume a range of plant, algae,animal and detrital material. What they consume dependson availability and developmental stage of the animal. Forexample, Cherax destructor, which have been introducedinto rivers in Western Australia, consume mainly adultmosquito Wsh (Gambusia holbrooki) in summer and mainlyplant material in winter (Beatty 2006). These feeding habitsreXect the availability of the dietary items. The diet ofC. destructor in Lake Eucumbene, NSW, Australia con-sisted of mainly plant material (80% seeds and grasses) andminor amounts of arthropods (chironomid larvae and pupaeand crayWsh) (Faragher 1983).

Lipids and fatty acids

The major storage forms of lipids in crustaceans are triacyl-glycerides. Additionally, wax esters appear in some spe-cies. Triacylglycerides are preferentially used to covershort-term energy demands e.g. in marine copepods,whereas wax esters serve as long-term energy stores(Hagen and Auel 2001). Although low, E. sericatus showedsigniWcantly higher amounts of wax esters than C. destruc-tor. Accordingly, E. sericatus may be better suited to over-come periods of starvation, which is in accordance with itslife style and its preference for ephemeral habitats.

Some of the fatty acids, which were components of thetriacylglycerols, were mostly likely synthesised endoge-nously by the crayWsh and hence cannot provide reliableinformation on dietary preferences. Palmitic acid (16:0) andstearic acid (18:0), both of which were present in largeamounts in the midgut gland of both species, are typicalproducts of fatty acid biosynthesis in both plants and ani-mals. Desaturation and elongation of palmitic acid (16:0) toproduce palmitoleate (16:1) and oleate (18:1) are known toendogenously occur in copepods and may also take place inthe midgut gland of the crayWsh. In contrast the poly unsat-urated fatty acids, linoleate (18:2(�-6)) and linolenate(18:3(�-3)) are synthesised by plants and not animals andare indicative of the consumption of terrestrial plant mate-rial in coastal marine ecosystems (Budge et al. 2001). In the

123

J Comp Physiol B

midgut glands of both species, these fatty acids separatelyaccounted for more than 15% and together accounted formore than 40% of total fatty acids. This suggests that bothC. destructor and E. sericatus consumed substantialamounts of terrestrial plants. However, the amounts of eachfatty acid were signiWcantly higher in C. destructor than inE. sericatus indicating that C. destructor may have ingestedand assimilated more plant material than E. sericatus.

Similarity analysis revealed that the fatty acid composi-tion not only distinctly diVered between plants andcrayWshes but also between both crayWsh species. Althoughthe crayWshes obviously accumulated typical plant fattyacids [18:2(�-6) and 18:3(�-3)], many other fatty acids didnot match with the pattern of the analysed plants due toendogenous conversion of fatty acids or due to the accumu-lation of fatty acids from other food sources than those ana-lysed. Particularly, smaller invertebrates including insectsand annelids form staple dietary items as shown by gutcontent analysis. The diVerent fatty acid patterns betweenC. destructor and E. sericatus are certainly due to diVer-ences and quantities of diets but may also be related todiVerences of metabolic pathways of either species.

Digestive enzymes

Both species are able to hydrolyse cellulose associated withthe cell walls of plants. However, E. sericatus seems to bemore eYcient at cellulose hydrolysis than C. destructorbecause the midgut gland of the former species showshigher activities of total cellulase and the enzymes contrib-uting to this, endo-�-1,4-glucanase and �-glucosidase.E. sericatus is thus better able to hydrolyse, digest and uti-lise cellulose than C. destructor. Since both species areclosely related, the higher total cellulase, endo-�-1,4-glu-canase and �-1,4-glucosidase activity may indicate thatE. sericatus has adapted to a diet containing more terrestrialplant material.

Hemicellulases, laminarinase, lichenase and xylanasewere present at similar levels in the midgut glands of bothE. sericatus and C. destructor. Xylanase activity is indica-tive of the ability to hydrolyse xylan, a common structuralpolysaccaride present in the cell walls of grasses and otherangiosperms (Bacic et al. 1988). Lichenase activity indi-cates that both species are able to hydrolyse lichenan ormixed linkage �-glucans. Lichenan or mixed linkage �-glu-can is composed of glucose units joined by mainly �-1,4glycosidic bonds with some �-1,3 linkages and is a majorcomponent of the cell walls of cereals and grasses(McCleary 1988; Terra and Ferreira 1994). Xylanase andlichenase activities, along with the total cellulase activityindicate that both species are capable of digesting the majorcomponents of the cell walls of grasses, which may becommonly found in their respective habitats. Laminarianse

activity indicates that both E. sericatus and C. desctructorare able to hydrolyse laminarin. Laminarin is a principally�-1,3 glycosidic bond glucose polymer (Terra and Ferreira1994). It is a common storage polysaccharide of algae butalso present in phloem and plant wound tissue (Vonk andWestern 1984; Terra and Ferreira 1994). Thus, both speciesmay be able to digest algae. E. sericatus feeds on rootsgrowing into their feeding chambers. These roots maycontain substantial amounts of wound tissue and hencelaminarin.

The ability to digest the plant compounds cellulose andhemicellulose is widespread amongst the Crustacea(reviewed by Linton and Greenaway 2007). The presenceof xylanase and lichenase, however, has recieved scantattention. Like laminarinase, these enzymes may be widelydistributed throughout the Crustacea. However they mayonly be present in species that encounter these substrates intheir diet (Crawford et al. 2005).

Despite contradicting reports published recently (Pavasovicet al. 2006), the cellulases and hemicellulases are mostlikely responsible for the digestion of dietary Wbre. Assim-ilation co-eYcients for crude Wbre by Cherax destructorfed an artiWcial diet were high (Jones and De Silva 1997)indicating that the activity of cellulase enzymes mayaccount for such cellulose digestion. Similarly highercellulose assimilation co-eYcients for the gecarcinidcrabs, Gecarcoidea natalis and Discoplax hirtipes fedbrown leaves is correlated with higher total cellulase activ-ities in these species (Greenaway and Linton 1995; Lintonand Greenaway 2004). In addition, high hemicelluloseassimilation co-eYcients for G. natalis and D. hirtipes feda brown leaf litter diet correlated well with substantialactivities of the hemicellulase enzymes, laminarinase,lichenases and xylanase (Linton and Greenaway 2004).

C. destructor and E. sericatus both possess proteinasesfor hydrolysing protein associated with plant and animalmaterial. Higher total protease activity in C. destructormay indicate that this species has a higher intake of animalmaterial than E. sericatus. Like decapods generally, theproteases present in the midgut glands of both C. destructorand E. sericatus were mainly serine proteinases (trypsin-and chymotrypsin-like enzymes) with a small amounts ofcysteine proteinases (Ceccaldi 1997; Johnston andYellowlees 1998; Lehnert and Johnson 2002; Navarrete delToro et al. 2006; Gudmundsdottir 2002). The proteolyticactivity remaining after the inhibition of both serine andcysteine proteinases (20% in E. sericatus and 15% inC. destructor) may represent activity of aspartyl proteasesand metalloproteases such as carboxypeptidase A andcarboxypeptidase B.

Both E. sericatus and C. destructor possessed substan-tial activities of endochitinase and N-acetyl-�-D-glucosa-minidase. Both enzymes work in concert to hydrolyse

123

J Comp Physiol B

chitin to its monomer of N-acetyl-glucosamine. Endoch-itinase hydrolyses the internal glycosidic bonds of chitinto produce dimers and trimers of N-acetyl-D-glucosamine(Kono et al. 1990; Genta et al. 2006). N-acetyl-�-D-glu-cosaminidase then hydrolyses the oligomers intomonomers (Peters et al. 1999; Zou and Fingerman 1999;Xie et al. 2004). Chitinase enzymes are ubiquitousthroughout the Crustacea acting as both a moulting anddigestive enzyme (Zou and Fingerman 1999; Peters et al.1999). C. destructor may be more eYcient at digestingarthropod chitin given its midgut gland possessed ahigher activity of N-acetyl-�-D-glucosaminidase than themidgut gland of E. sericatus. This, plus the higher totalprotease activity within the midgut gland suggests thatC. destructor is better suited to digest animal tissue, e.g.from arthropods than E. sericatus.

The presence of proteinases, endo-�-1,4-glucanase,�-1,4-glucanase, xylanase, lichenase and, laminarinase hasbeen documented previously in Cherax species (Crawfordet al. 2005; Figueiredo et al. 2001). CrayWsh of the genusCherax are omnivorous consuming mainly plant materialand clearly their digestive enzyme complement correlateswell with the substrates that they would encounter in theirnatural diet.

Conclusion

Both C. destructor and E. sericatus are omnivorous, con-suming a range of plant, animal and algal material. Thecomplement of digestive enzymes (proteinases, laminarin-ase, lichenase, xylanase, total cellulase, endo-�-1,4-gluco-sidase, �-1,4-glucosidase) within the midgut gland suggeststhat both species are capable of eYciently hydrolysing thewide variety of substrates associated with such a diet.E. sericatus seems to be better adapted to cope with terres-trial plant material than C. destructor given its midgutgland possess higher activities of total cellulase, endo-�-1,4-glucanase and �-1,4-glucosidase and the morphology ofthe gastric mill suggests that it is better able to cut Wbrousplant material. In contrast, the morphology of the gastricmill and the higher total cellulase and N-acetyl-�-D-glu-coasminidase activities possessed by C. destructor suggeststhat this species is better able to digest animal material inthe form of arthropods. Given that E. sericatus andC. destructor are closely related (Crandall et al. 1999), thediVerences in enzyme activity and morphology of thegastric mill observed between the two species may repre-sent E. sericatus being better adapted to digest terrestrialplant material. The contents of the faeces suggest that whileboth species consumes large amounts of plant material,C. destructor consumes signiWcantly more aquatic macro-algae.

Acknowledgments The International OYce of the German Ministryof Education and Research supported this work by a travel grant toR. Saborowski (AUS 02/06A). The Hermon Slade Foundationprovided a research grant to S. Linton.

References

Bacic A, Harris PJ, Stone BA (1988) Structure and function of plantcell wall. In: Stumpf PK, Conn EE (eds) The Biochemistry ofPlants Vol 14. Academic Press, New York, pp 297–371

Beatty SJ (2006) The diet and trophic positions of translocated,sympatric populations of Cherax destructor and Cherax cainii inthe Hutt River, Western Australia: evidence of resource overlap.Mar Freshw Res 57:825–835

Bradford MM (1976) A rapid and sensitive method for the quantitationof microgram quantities of protein utilizing the principle ofprotein-dye binding. Anal Biochem 72:248–254

Budge SM, Parrish CC, Mckenzie CH (2001) Fatty acid compositionof phytoplankton, settling particulate matter and sediments at asheltered bivalve aquaculture site. Mar Chem 76:285–303

Ceccaldi HJ (1997) Anatomy and physiology of the digestive system.In: D’Abramo LR, Conklin DE, Akiyama DM (eds) Advances inworld aquaculture, vol. 6, Crustacean nutrition. The worldaquaculture society, Louisiana state university, Baton Rouge, pp261–291

Crandall KA, Fetzner JW, Lawler SH, Kinnersley M, Austin CM(1999) Phylogenetic relationships among the Australian and NewZealand genera of freshwater crayWshes (Decapoda: Parastaci-dae). Aust J Zool 47:199–214

Crawford AC, Richardson NR, Mather PB (2005) A comparative studyof cellulase and xylanase activity in freshwater crayWsh and ma-rine prawns. Aquac Res 36:586–592

Dalsgaard J, St John M, Kattner G, Müller-Navarra D, Hagen W(2003) Fatty acid trophic markers in the pelagic marine environ-ment. Adv Mar Biol 46:225–340

Erlanger BF, Kokowsky N, Cohen W (1961) The preparation andproperties of two new chromogenic substrates for trypsin. ArchBiochem Biophys 95:271–278

Faragher RA (1983) Role of the crayWsh Cherax destructor Clark asfood for trout in Lake Eucumbene, New South Wales. Aust J MarFreshw Res 34:407–417

Figueiredo MSRB, Kricker JA, Anderson AJ (2001) Digestive enzymeactivities in the alimentary tract of the redclaw crayWsh, Cheraxquadricarinatus (Decapoda: Parastacidae). J Crust Biol 21:334–344

Fraser AJ, Tocher DR, Sargent JR (1985) Thin-layer chromatography-Xame ionization detection and the quantitation of marine neutrallipids and phospholipids. J Exp Mar Biol Ecol 88:91–100

Genta F, Blanes L, Cristofoletti PT, do Lago CL, Terra WR, Ferreira C(2006) PuriWcation, characterisation and molecular cloning of themajor chitinase from Tenebrio molitor larval midgut. Insect Bio-chem Mol Biol 36:789–800

Giddins RL, Lucas JS, Neilson MJ, Richards GN (1986) Feeding ecol-ogy of the mangrove crab Neosarmatium smithi (Crustacea:Decapoda: Sesarmidae). Mar Ecol Prog Ser 33:147–155

Goddard JS (1988) Food and Feeding. In: Holdich DM, Lowery RS(eds) Freshwater crayWsh: biology management and exploitation.Croom Helm, London, pp 145–166

Greenaway P, Linton SM (1995) Dietary assimilation and food reten-tion time in the herbivorous terrestrial crab Gecarcoidea natalis.Physiol Zool 68:1006–1028

Growns IO, Richardson AMM (1988) Diet and burrowing habits of thefreshwater crayWsh, Parastacoides tasmanicus tasmanicus Clark(Decapoda, Parastacidae). Aust J Mar Freshw Res 39:525–534

123

J Comp Physiol B

Gudmundsdottir A (2002) Cold-adapted and mesophilic brachyurins.Biol Chem 383:1125–1131

Hagen W (2000) Lipids. In: Harris RP, Wiebe PH, Lenz J, SkjoldalHR, Huntley M (eds) ICES zooplankton methodology manual.Academic Press, San Diego, pp 113–119

Hagen W, Auel H (2001) Seasonal adaptations and the role of lipids inoceanic zooplankton. Zoology 104:313–326

Heeren T, Mitchell BD (1997) Morphology of the mouthparts, gastricmill and digestive tract of the gian crab, Pseudocarcinus gigas(Milne Edwards) (Decapoda: Oziidae). Mar Freshw Res 48:7–18

Heinzel H-G (1988) Gastric mill activity in the Lobster I. Spontaneousmodes of chewing. J Neurophysiol 59:528–550

Heinzel H-G, Weinmann JM, Marder E (1993) The Behavioural reper-toire of the gastric mill in the crab, Cancer pagurus: an in situendoscopic and electrophysiological examination. J Neurosci13:1793–1803

Horwitz P (1990) A taxonomic revision of species in the freshwatercrayWsh genus Engaeus Erichson (Decapoda: Parastacidae). In-vertebTaxon 4:427–614

Johnston D, Freeman J (2005) Dietary preference and digestive en-zymes activities as indicators of trophic resourece utilization bysix species of crab. Biol Bull 208:36–46

Johnston DJ, Yellowlees D (1998) Relationship between dietary pref-erences and digestive enzyme complement of the slipper lobsterThenus orientalis (Decapoda: Scyllaridae). J Crust Biol 18:656–665

Jones PL, De Silva SS (1997) Apparent nutrient digestibility of formu-lated diets by the Australian freshwater crayWsh Cherax destruc-tor Clark (Decapoda, Parastacidae). Aquac Res 28:881–891

Jones D, Morgan G (1994) A Weld guide to Crustaceans of AustralianWaters. Reed, Sydney

Kattner G, Fricke HSG (1986) Simple gas-liquid chromatographymethod for the simultaneous determination of fatty acids andalcohols in wax esters of marine organisms. J Chromatogr361:263–268

Kono M, Matusi T, Shimizu C, Koga D (1990) PuriWcations and someproperties of chitinase from the liver of a prawn, Penaeus japoni-cus. Agr Biol Chem 54:2145–2147

Kunze J, Anderson DT (1979) Functional morphology of the mouth-parts and gastric mill in the hermit crabs Clibanarius taeniatus(Milne Edwards), Clibanarius virescens (Krauss), Paguristessquamosus (McCulloch) and Dardanus setifer (Milne-Edwards)(Anomura: Paguridae). Aust J Mar Freshw Res 30:683–722

Lehnert SA, Johnson SE (2002) Expression of hemocyanin and diges-tive enzyme messenger RNAs in the hepatopancreas of the BlackTiger Shrimp, Penaues monodon. Comp Biochem Physiol B133:163–171

Linton SM, Greenaway P (2004) Presence and properties of cellulaseand hemicellulase enzymes of the gecarcinid land crabsGecarcoidea natalis and Discoplax hirtipes. J Exp Biol207:4095–4104

Linton SM, Greenaway P (2007) A review of feeding and nutrition ofherbivorous land crabs: adaptations to low quality plant diets.J Comp Physiol B 177:269–286

Martin JW, Jourharzadeh P, Fitterer PH (1998) Description andcomparison of major foregut ossicles in hydrothermal vent crabs.Mar Biol 131:259–267

McCleary BV (1988) Licheninase from Bacillus subtilis. In: WoodWA, Kellogg ST (eds) Methods in Enzymology, vol 160.Biomass Part A. Cellulose and Hemicellulose. Academic Press,New York, pp 572–575

Merrick JR (1993) Freshwater crayWshes of New South Wales. Lin-nean society of New South Wales, Sydney

Navarrete del Toro MDLA, Garcia-Carreno FL, Lopez MD, Celis-Guerrero L, Saborowski R (2006) Aspartic proteinases in thedigestive tract of marine decapod crustaceans. J Exp Zool305A:645–654

Nelson DL, Cox MM (2005) Lehninger Principles of Biochemistry.Freeman and Company, New York, p 1119

Pavasovic A, Richardson NA, Mather PB, Anderson AJ (2006) InXu-ence of insoluble dietary cellulose on digestive enzyme activity,feed digestibility and survival in the red claw crayWsh, Cheraxquadricarinatus (von Martens). Aquac Res 37:25–32

Peters G, Saborowski R, Buchholz F, Mentlein R (1999) Two distinctforms of the chitin-degrading enzyme N-acetyl-�-D-glucosamini-dase in the Antarctic krill: specialists in digestion and moult. MarBiol 134:697–703

Saborowski R, Buchholz F, Vetter R-A H, Wirth S, Wolf G (1993) Asoluble dye-labelled chitin derivative adapted for the assay of krillchitinase. Comp Biochem Physiol B 105:673–678

Saborowski R, Sahling G, Navarette del Toro MA, Waltera I, García-Carreñob FL (2004) Stability and eVects of organic solvents onendopeptidases from the gastric Xuid of the marine crab Cancerpagurus. J Mol Catal B: Enzymatic 30:109–118

Salindeho IR, Johnston DJ (2003) Functional morphology of themouthparts and proventriculus of the rock crab Nectocarcinus tu-berculosus (Decapoda: Purtunidae). J Mar Biol Assoc UK83:821–834

Scrivener AM, Slaytor M (1994) Properties of the endogenous cellu-lase from Panesthia cribrata Saussure and puriWcation of majorendo-�-1, 4-glucanase components. Insect Biochem Mol Biol24:223–231

Skilleter GA, Anderson DT (1986) Functional morphology of thechelipeds, mouthparts and gastric mill of Ozius truncatus (MilneEdwards) (Xanthidae) and Leptograpsus variegatus (Fabricius)(Grapsidae) (Brachyura). Aust J Mar Freshw Res 37:67–79

Suter PJ, Richardson AMM (1977) The biology of two species of Enga-eus (Decapoda: Parastacidae) in Tasmania III.Habitat, food, associ-ated fauna and distribution. Aust J Mar Freshw Res 28:95–103

Suthers IM, Anderson AJ (1981) Functional morphology of mouth-parts and gastric mill of Ibacus peronii (Leach) (Palinura: Scyllar-idae). Aust J Mar Freshw Res 32:931–944

Terra WR, Ferreira C (1994) Insect digestive enzymes: properties,compartmentalization and function. Comp Biochem Physiol109B:1–62

Teschke M, Saborowski R (2005) Cysteine proteinases substitute forserine proteinases in the midgut gland of Crangon crangon andCrangon allmani (Decapoda, caridea). J Exp Mar Biol Ecol316:213–229

Vonk HJ, Western JRH (1984) Comparative biochemistry and physi-ology of enzymatic digestion. Academic Press, New York

Watanabe H, Nakamura M, Tokuda G, Yamaoka I, Scrivener AM,Noda H (1997) Site of secretion and properties of endogenousendo-�-1, 4-glucanase components from Reticulitermes speratus(Kolbe), a Japanese subterranean termite. Insect Biochem MolBiol 27:305–313

Xie X-L, Chen Q-X, Lin J-C (2004) PuriWcation and some propertiesof �-N-acetyl-D-glucosaminidase from prawn (Penaeus vanna-mei). Mar Biol 146:143–148

Zou E, Fingerman M (1999) Chitobiase activity in the epidermis andhepatopancreas of the Wddler crab Uca pugilator during themolting cycle. Mar Biol 133:97–101

123